Write an effective prompt
The R.C.T.F.C. method and three habits to get a useful answer by the second exchange.
In this guide
The R.C.T.F.C. method and three habits to get a useful answer by the second exchange.
Leadership, HR, sales, finance, order management, procurement, quality: templates you can adapt in a minute.
Five rules to protect your data and stay in control of what leaves the company.
Share a common library, then spot the tasks that deserve automation.
01
An effective prompt has five parts: a role (who the AI should be), a context (the situation), a precise task, an output format and constraints. Remember them as R.C.T.F.C. Three habits then make the difference: show an example of what you expect, ask the AI to question you before answering, and iterate, because the first answer is only a draft.
1
Show an example
A sample of what you expect is worth more than ten lines of description.
2
Let it ask questions
Add: “Before answering, ask me the questions you need to do this well.”
3
Iterate
The first answer is a draft. Refine it: “shorter”, “more direct”, “add a table”.
Before
Write an email to chase a customer.
After
You are an order management lead. Write a reminder to a loyal industrial customer whose €12,400 invoice is 30 days overdue. Friendly but firm, six lines maximum, offer a payment plan if needed.
02
Executives spend much of their time reading, arbitrating and communicating. AI helps prepare a structured leadership meeting, turn a fifty-page report into one useful page, stress-test a strategic decision by playing devil’s advocate, and write a sensitive internal announcement in the right tone.
Here is [a report / an audit / a market study] of [number] pages. Summarise it for a busy executive: 1. The 3 key messages, one sentence each 2. The figures to remember 3. Risks and opportunities for our company ([describe the business]) 4. The 3 questions to ask the author Format: one page maximum.
Replace the items in [brackets] with your own information.
+ 3 more prompts in the PDF guide03
From hiring to annual reviews, HR produces a lot of standardised writing. AI speeds up attractive, inclusive job ads, structured interview grids, balanced feedback preparation and a week-by-week onboarding plan.
You are the recruitment lead at a [SME/mid-size company] in [industry] based in [city]. Write a job ad for a [job title] position. Main responsibilities: [list] Profile: [skills, experience] What we offer: [salary, remote work, benefits, culture] Constraints: inclusive language, no clichés (“dynamic”, “passionate”), 400 words maximum, an opening line that makes people read on.
Replace the items in [brackets] with your own information.
+ 3 more prompts in the PDF guide04
Preparing meetings, prospecting, writing proposals: AI saves hours on each of these every week. This chapter covers preparing a prospect meeting, outbound sequences, turning meeting notes into a proposal, repurposing content across channels and learning from a lost deal.
I am meeting [company], [industry], [size], on [date]. Contact: [role]. Our offer: [short description]. Prepare: - 3 likely challenges for this type of company right now - 5 open discovery questions - the 3 most likely objections and an answer to each - a next step to agree on at the end of the meeting
Replace the items in [brackets] with your own information.
+ 4 more prompts in the PDF guide05
Finance handles numbers but spends a lot of time explaining them. AI helps analyse a budget variance and write the management commentary, build a complex Excel formula, prioritise collections, explain financial statements to non-finance managers and comment on a cash forecast.
You are a financial controller. Here are actuals vs budget for [period]: [paste table]. 1. Identify the 5 most significant variances (in value and %) 2. Suggest possible explanations for each 3. List the questions to ask operations to confirm 4. Write a 10-line management commentary for leadership
Replace the items in [brackets] with your own information.
+ 4 more prompts in the PDF guide06
Order management and customer service handle a continuous flow of requests. AI helps answer a complaint without deflecting, sort a shared inbox, announce a delivery delay and build an FAQ from recurring requests, surfacing process problems along the way.
You are the customer service lead at a [industry] company. Here is the complaint: [paste message]. Internal facts: [delay, error, responsibility...]. Write a reply that: - acknowledges the problem without deflecting - explains what happened in 2 sentences - offers a concrete solution and timeframe - preserves the business relationship Tone: professional and empathetic, 150 words maximum.
Replace the items in [brackets] with your own information.
+ 3 more prompts in the PDF guide07
Buyers compare, negotiate and secure supply. AI builds a comparison table of supplier offers, prepares a negotiation (target, walk-away point, trade-offs), helps write a complete specification and assesses supplier dependency risk with a mitigation plan.
You are a buyer at a [company/industry]. Here are [number] offers for [product/service]: [paste or attach offers]. 1. Build a comparison table: price, lead times, payment terms, warranties, penalties, points of attention 2. Identify missing or ambiguous items in each offer 3. Recommend a supplier with reasons, and suggest 3 negotiation levers
Replace the items in [brackets] with your own information.
+ 3 more prompts in the PDF guide08
Quality relies on proven methods that AI can apply quickly. This chapter covers non-conformity analysis (5 whys, Ishikawa), writing a procedure a newcomer can follow, preparing an audit and running a blame-free lessons-learned session.
You are a quality manager in [industry]. Non-conformity observed: [description, frequency, impact]. Run a root-cause analysis with the 5 whys, then an Ishikawa (5M) diagram presented as a list. Then propose corrective and preventive actions, with owner and typical timeframe.
Replace the items in [brackets] with your own information.
+ 3 more prompts in the PDF guide09
Generative AI is a powerful tool that calls for a few simple rules. They protect your data, your company and the quality of your work.
Employee or customer personal data, unpublished financial information, trade secrets: use a business plan (Claude Team or Enterprise, ChatGPT Enterprise, Copilot M365) whose terms exclude training on your data.
AI can make up figures, references or legal articles. Anything sent outside the company must be reviewed.
AI prepares; people decide and sign.
Replace names with roles (“customer A”, “employee X”) whenever possible.
An AI usage policy protects the company and reassures teams.
10
A good prompt saves one person time. A shared library saves the whole company time. Four steps take you from individual use to shared practice, and then help you spot the tasks worth automating end to end.
01
Gather the prompts that work in a shared space: Claude Projects, a shared workspace or a simple document.
02
Describe the company context once (business, tone, customers, vocabulary) in reusable instructions.
03
One go-to person per department grows the library and helps colleagues.
04
Quantify the time saved on 3 or 4 recurring tasks to spot candidates for automation.
If a task comes back every day and always follows the same pattern (invoice processing, sorting order-desk emails, comparing offers), it deserves better than copy-and-paste: it can be automated end to end, with an AI agent connected to your tools and human review where it matters.
Automate your processesFree guide
Enter your work email and the PDF downloads straight away.
The PDF includes
FAQ
We find the repetitive tasks that deserve better than copy-and-paste, and automate them with your tools.
A 30-minute first call, no commitment.